Phalanx or legion: what does the evidence say?
The Macedonian pike phalanx and the Roman manipular legion met in a series of battles between 280 and 168 BC. The phalanx side won the first two, under Pyrrhus at Heraclea and Asculum, and the legions won at Cynoscephalae in 197, Magnesia in 190 and Pydna in 168 BC. Historians have not settled which system was better. Polybius granted the phalanx its frontal power and explained its defeats by the ground, the turned flank and the gaps that opened in pursuit. Mommsen blamed command and cavalry as much as the formation. This page sets out the evidence for each and leaves the final verdict to you.
The phalanx here is the Macedonian formation of Philip II and Alexander and their successors: men packed deep in files of up to sixteen, with very long pikes. The legion is the Roman army of the third and second centuries BC, built of small companies called maniples. Treccani says Philip II perfected the phalanx and describes the legion's three lines of maniples, and Asclepiodotus sets out how the tacticians built the phalanx file by file.1
This is not a ranking. Below are the two systems side by side, every battle in which they met, and six questions on which a fighting formation can be judged, each with the strongest evidence for the phalanx and for the legion. At the end the decision is yours.
Side by side
- Period compared
- From Philip II in the 4th century BC to Pydna, 168 BC
- The manipular legion of the 3rd and 2nd centuries BC
- Main weapon
- The sarissa, 16 cubits long by the original design, 14 in practice
- Two javelins and a sword good for cutting and thrusting
- Formation
- Sixteen deep, three feet a man, five pikes beyond each man in front
- Three lines of maniples with gaps, each man with room to fight
- Unit
- Syntagma of 256 men, 16 by 16
- Legion of 4,200 foot and 300 horse, in maniples
- Soldiers
- The Macedonian levy, the flower of the kingdom
- Citizens sorted by age and wealth
- Ground it needs
- Level and clear
- Can fight in detachments or singly
- Battles where they met
- Heraclea 280 and Asculum 279 won by the phalanx side, and Perseus' early win over the consul Licinius
- Cynoscephalae 197, Magnesia 190, Pydna 168 won by the legion side
Weapons, formation and recruitment as described by Polybius, Asclepiodotus and the Treccani encyclopedia. The images are a 19th-century drawing of a syntagma and a later imperial relief, not portraits of these armies. We give no win and loss totals beyond the named battles.2
Where they met: from Heraclea to Pydna
Rome had good reason to fear the formation. Justin says that when the Senate weighed war with Philip V, the Romans remembered how much Pyrrhus had shaken Italy with a small band of Macedonians, and what the Macedonians had done in the East.3
Pyrrhus in Italy, 280 to 279 BC
The first meetings went to the phalanx side. Pyrrhus of Epirus, Dionysius of Halicarnassus writes, put the Macedonian phalanx on his right wing at Asculum, with Italian and Greek phalanxes beside it and his elephants behind both wings, against two consular armies. Plutarch says over fifteen thousand men fell on both sides, and quotes Pyrrhus' answer to a man who congratulated him: one more victory over the Romans and we are utterly ruined.4
Florus describes the elephants at Heraclea, which so frightened the Roman horses with their bulk, their smell and their trumpeting that they caused panic and destruction far and wide.5
Cynoscephalae, 197 BC
Philip V met the consul Flamininus in the hills of Thessaly with sixteen thousand phalangites, the flower of his kingdom, Livy says, and the Roman army almost equal in numbers. The armies blundered into each other in mist. Philip's right wing, charging downhill, drove the Roman left back, and Mommsen calls the attack of the phalanx irresistible on so favourable ground.6
His left wing was still coming up over the crest. Roman elephants scattered it, and then, Polybius writes, a tribune with not more than twenty maniples wheeled round and fell on the victorious Macedonian right from behind. Since the phalanx cannot turn about or fight man to man, it broke. Polybius gives about seven hundred Roman dead against eight thousand Macedonians killed and five thousand taken.7

Magnesia, 190 BC
Antiochus III's army included a phalanx with elephants between its divisions. When his left wing broke, Livy writes, the fugitives ran back into the phalanx, and its men could not use their extraordinarily long spears, which the Macedonians call sarisae. Mommsen adds that the Romans took care not to attack the phalanx with their legions, sending archers and slingers against the packed mass instead, and that it retired in good order until its own elephants panicked.8
Eutropius gives 50,000 of the king's infantry and 3,000 horse killed in that battle, near Mount Sipylus by Magnesia.9
Pydna, 168 BC
Perseus, Philip's son, began the war well: Eutropius records that he beat the consul Licinius in a heavy battle. At Pydna the consul Aemilius Paullus faced him on the plain, and Plutarch says that on seeing the phalanx advance Aemilius felt he had never seen a sight more fearful.10
The Paeligni charged first and were cut down. As the phalanx followed across uneven ground, gaps opened in it, and Plutarch says Aemilius ordered his cohorts to plunge into the empty spaces and fight not one battle but many separate ones. The battle was over within an hour. Plutarch gives over 25,000 Macedonian dead and 80 or 100 Romans, Eutropius 20,000 Macedonian foot killed and 100 Romans.11

Frontal power
The case for the phalanx
Polybius, the legion's great advocate, concedes the point at the start of his famous comparison: when the phalanx has its characteristic virtue and strength, nothing can sustain its frontal attack or withstand its charge. Each man in the front rank has the points of five pikes beyond him, and a single Roman must face two phalangites and so ten pikes, which he cannot cut through in time.12
Livy agrees. At Pydna, he writes, as long as the phalanx was compact with its front bristling with levelled spears, its strength was irresistible, and if the whole Roman line had charged it while unbroken they would have spitted themselves on its spears. Earlier, describing Philip's army, he writes that on favourable ground and in a fair field the Macedonian phalanx would always remain unbeaten.13
The Romans themselves fought shy of it. Frontinus says the consul Manius Curius, seeing that the phalanx of King Pyrrhus could not be resisted when it was drawn out in line, took care to fight in confined places where it would be crowded together and get in its own way.14
The case for the legion
The legion did not need to win the frontal fight. Polybius asks why, if the phalanx is so strong, the Romans win, and answers that in war the time and place of action are uncertain, while the phalanx has only one time and one place in which it can do its work. Livy says the legions at Pydna won by many separate engagements that first shook the phalanx out of formation and then broke it.15
Frontinus tells another version of Pydna in which Aemilius sent the cavalry of his left wing at full speed across the front of the phalanx, covering themselves with their shields, so that the enemy's spear points broke on them, and the Macedonians, deprived of their spears, broke and fled.16
Terrain and flexibility
The case for the phalanx
On the right ground the phalanx did what nothing else could. At Cynoscephalae Philip's right wing, charging downhill, broke the Roman left. At Magnesia the phalanx retired in good order under missile fire. Mommsen notes that the Macedonians kept their phalanx, often conquered, because they still believed it invincible.17
It was not meant to be rigid. Treccani says Philip II perfected the phalanx, giving it the greatest possible shock while making it far more mobile and well protected on its flanks and rear.18
Its defeats came when it was caught out of order: at Cynoscephalae the left was still in column, Livy writes, marching rather than fighting, when the Romans struck, and at Pydna, Plutarch says, the phalanx broke only after its advance across uneven ground had opened gaps.19
The case for the legion
Polybius lists what the phalanx needs: level and clear ground with no ditches, clefts, clumps of trees, ridges or water courses. The phalanx soldier can serve neither in detachments nor singly, he writes, while every Roman soldier, once armed, can adapt himself to every place and time and meet attack from every quarter.20
The legion was built for that. Treccani describes three separate lines of heavy infantry, 1,200 hastati, 1,200 principes and 600 triarii, with 1,200 light velites to open the fight, grouped in maniples of 120 men, or 60 in the third line, and 300 horsemen on the wings.21
Livy explains why a turned flank was fatal: forced to bring round their long, heavy spears, the phalangites became a confused mass, and if attacked suddenly on the flank or rear they went to pieces like a falling house. At Cynoscephalae, he adds, the heavy and immobile formation could not face round on a new front.22
Combined arms and command
The case for the phalanx
The phalanx was never meant to fight alone. Pyrrhus' victories came with elephants and cavalry on the wings, as Dionysius' battle order at Asculum shows. At the Hydaspes, Arrian writes, Alexander's phalanx advanced against Porus' elephants, its men casting darts at the riders and striking the beasts, and when the tired elephants backed away it locked shields and pressed in from all sides.23
The theorists planned for it too. Asclepiodotus says most tacticians made the ideal heavy phalanx 16,384 men, a number that halves cleanly down to one, with a light-armed phalanx half that size beside it.24
Its later defeats often began elsewhere. At Magnesia it broke only after the left wing gave way, and at Pydna, Eutropius writes, the cavalry fled with the king, untouched. Mommsen writes that the Macedonian cavalry, which alone could have helped, looked calmly on and fled, the king among the first.25
The case for the legion
The Romans used every arm against the phalanx. At Cynoscephalae, Mommsen writes, elephants on the Roman right broke the Macedonian left before a resolute officer with twenty companies struck the victorious wing from behind, and at Magnesia archers and slingers did the work while the legions stood by. At Pydna it was the decision of one general, Plutarch writes, to break his line into many small attacks.26
Florus thought Cynoscephalae hardly deserved the name of a regular battle, once the Romans had forced their way into Macedonia through the gorges of the Aous.27
Manpower
The case for the phalanx
Macedon could field a large phalanx when it mattered. Perseus' 43,000 men, nearly half in the phalanx, were the largest Macedonian army since Alexander's, Livy says.28
Justin adds that Perseus had, besides a Macedonian army with a reputation for being unbeatable, ten years of war funds stored up by his father, and that he won the first cavalry clash of the war.29
It was also expensive. In a recent study Christopher Matthew reckons, as Michael J. Taylor sums him up, that Alexander's phalangites were paid four obols a day, with picked troops and officers paid more, a cost covered only once the Persian bullion was captured.30
The case for the legion
Rome could lose and refill. In 225 BC, Polybius says, the Romans and their allies able to bear arms numbered more than seven hundred thousand foot and seventy thousand horse. Each legion of 4,200 foot was drawn from citizens sorted by age and wealth, the youngest and poorest as light troops, the oldest as triarii in the third line.31
Pyrrhus saw the result at first hand. Plutarch says that after Asculum he had no more men to summon from home while his Italian allies were losing interest, and the Romans refilled their army as if from a fountain gushing forth indoors.32
Individual Romans showed the same spirit. Justin tells how at Pydna the son of Cato, unhorsed among the thickest of the enemy, lost his sword, fought his way back into the Macedonian ranks behind his shield to recover it, and how the others, copying his daring, won the victory.33
Losses and what they mean
The case for the phalanx
The lopsided casualty figures are ancient claims and should be read as such. At Cynoscephalae Livy follows Polybius' 8,000 Macedonian dead because he trusts him, and rejects Valerius Antias' 40,000 and Claudius Quadrigarius' 32,000 as exaggerations. Many Macedonians died after they had stopped fighting: Livy says they raised their pikes upright as a sign of surrender, and Roman soldiers who did not know the signal cut them down. Mommsen says the same.34
The case for the legion
Even with that caution the pattern holds across the battles: about 700 Romans against 8,000 Macedonians at Cynoscephalae in Polybius, 50,000 of Antiochus' infantry killed at Magnesia and 100 Romans against 20,000 Macedonians at Pydna in Eutropius.35
What came after
The case for the phalanx
The phalanx had built an empire before it met Rome. Alexander's phalanx crossed into Asia and won at the Granicus, Issus, Gaugamela and the Hydaspes, and Arrian credits him with great skill in marshalling and arming an army. Mommsen calls Pydna the last great battle of the phalanx, where 3,000 picked phalangites let themselves be cut down to the last man.36
The case for the legion
Pydna ended the Macedonian kingdom. Eutropius says all the cities of Macedonia surrendered and Perseus himself came into Aemilius' power, and Treccani names the battle as Aemilius' victory over the king. At Delphi Aemilius set up a monument whose frieze, the Delphi museum says, is the earliest relief illustrating a historical narrative, scenes from the battle of Pydna.37
The same museum keeps a marble head of Flamininus, who beat Philip V in 197 BC and was celebrated as the liberator of the Greek cities from Macedonian rule.38
How historians have judged them
The ancient judgement is mostly Polybius', and later writers built on it. He grants the phalanx its frontal power and explains its defeats by the ground, the turned flank and the gaps that opened in pursuit. Plutarch adds the testimony of Aemilius himself, who said he had never seen anything more fearful than the phalanx at Pydna.39
Mommsen put the defeats down to command and cavalry as much as to the formation: at Cynoscephalae Philip attacked before his left was formed, at Magnesia the legions were never engaged, and at Pydna the Macedonian horse looked on and fled. He wondered why Philip did not, like Hannibal, reorganise his army on the Roman model, and answered that the Macedonians still believed their phalanx invincible.40
The Delphi museum's description of the Aemilius Paullus monument shows how the winners chose to remember the meeting, as a historical scene carved in stone at Delphi. Florus, writing under the emperors, hardly counted Cynoscephalae as a battle at all.41
Battles on the site
Cynoscephalae and Pydna, the two battles the comparison rests on, now have their own pages. The others are the pages on each side, with the explainers around them.
Phalanx against legion · 197 BCBattle of Cynoscephalae
Phalanx against legion · 168 BCBattle of Pydna
ExplainerWhat Was a Phalanx?
ExplainerHow the Roman Legion Worked
The phalanx under Alexander · 333 BCBattle of Issus
The phalanx under Alexander · 331 BCBattle of Gaugamela
The phalanx against elephants · 326 BCBattle of the Hydaspes
Legions beaten in Africa · 255 BCBattle of Tunis
The legion's worst defeat · 216 BCBattle of Cannae
The legion against elephants · 202 BCBattle of Zama
CommanderPhilip II of Macedon
CommanderAlexander the Great
Guide · 336 to 323 BCThe Wars of Alexander the Great
GuideMilitary History of Ancient Rome
ExplainerWar Elephants in History
RankingThe Greatest Ancient Battles
See also Rome vs Carthage and Hannibal vs Alexander. Heraclea, Asculum and Magnesia have no page of their own yet.
Your verdict
We do not pick a winner. Here are the two cases at their strongest, and the questions that decide which one you find more convincing.
For the phalanx
- A front that, by Polybius' own account, nothing could withstand
- Heraclea and Asculum won against Roman armies
- Defeats that began on the wings or with the cavalry, not in the phalanx
- The army that conquered the Persian Empire under Alexander
- Casualty figures that come only from the winners
For the legion
- Cynoscephalae, Magnesia and Pydna, three decisive wins
- Men who could fight in detachments or singly, on any ground
- Small units that found the gaps in the phalanx
- A citizen levy that could replace any loss
- An end to the Macedonian kingdom within an hour at Pydna
Questions to weigh before you decide:
- Should a formation be judged on the ground it chose or on the ground it was given?
- Do Heraclea and Asculum count for the phalanx, or were they won by Pyrrhus' elephants and cavalry?
- Were Cynoscephalae and Pydna lost by the phalanx, or by Philip's haste and Perseus' cavalry?
- How much weight should we give Polybius, the main source for both sides?
- Is better about winning a single battle, or about who could replace losses for a generation?







